Large-scale neural networks implemented with non-volatile memory as the synaptic weight element: Comparative performance analysis (accuracy, speed, and power)
2015pp. 4.4.1–4.4.4
Citations Over TimeTop 10% of 2015 papers
Geoffrey W. Burr, Pritish Narayanan, R. M. Shelby, Severin Sidler, Irem Boybat, Carmelo di Nolfo, Yusuf Leblebici
Abstract
We review our work towards achieving competitive performance (classification accuracies) for on-chip machine learning (ML) of large-scale artificial neural networks (ANN) using Non-Volatile Memory (NVM)-based synapses, despite the inherent random and deterministic imperfections of such devices. We then show that such systems could potentially offer faster (up to 25×) and lower-power (from 120-2850×) ML training than GPU-based hardware.
Related Papers
- → Nonvolatile and volatile resistive switching characteristics in MoS2 thin film for RRAM application(2023)18 cited
- → Ferroelectric PZT Thin Films For Low-voltage Nonvolatile Memory(2005)4 cited
- Materials and Processes for Nonvolatile Memories II: Volume 997 2007 MRS Proceedings(2014)
- → Coexistence of Nonvolatile Unipolar Memory and Volatile Threshold Resistance Switching In LaMnO3 Thin Films(2020)
- Materials and processes for nonvolatile memories II : symposium held April 10-13, 2007, San Francisco, California, U.S.A.(2007)